GALIANO AND BLACK GLACIERS 83
observations, and in the ten years since the Boundary Survey photograph was made. Accordingly, although the full significance of the changes were not then clear to us, we made comparative observations and photographs from Russell's sites and studied the glacier and vicinity with his descriptions before us. There were two noteworthy changes, one in the glacier itself, the other in the alluvial fan just described.
On the glacier the alder and forest cover was entirely destroyed, and our 1905 photographs show barren ablation moraine from far up the mountain valley clear down to the visible front of the piedmont bulb. Even on the most stagnant eastern portion the forest was gone, and here were found quantities of dead wood, including mature tree trunks. In other places mature alder littered the surface of the piedmont bulb or was mixed with the morainic debris, and in two places the dead trunks were seen in considerable numbers, upright and in place. Near the centre of the glacier, 60 feet above sea level, we found twenty or more small trees, the oldest with 25 annual rings (PI. XXX, B). They increased in size upward, as if standing upside down, but had downward-pointing roots for the lower visible 8 or 4 feet. These roots extended from shaggy bark, proving that the upward-expanding tree remnants were not tap roots, and indicating that the last months of life of these trees involved such adverse circumstances of soil encroachment that roots were sent out at higher and higher levels as the lower portions were buried. A yard or less above the highest roots the trees were torn off and the frayed splinters were bent down.
The alluvial fan was as completely altered as the forest on the glacier. Not only could it no longer be seen from the sites of Russell's pictures, which we carefully located, but there was no such perfect fan anywhere between the Galiano Glacier and the Kwik River. In its place was a series of low, undulating moraine hills (PL XXX, A), some of them strewn with angular bowlders, such as characterize ablation moraine, between which flowed the muddy branches of Esker and Galiano streams. These streams were building up the depressions in the moraine and here and there were trimming the edges of the knolls. Besides the muddy glacial streams there were innumerable cold, clear-water streams emerging from the moraine, and small pools of clear, cold water in the depressions. There was a roughly concentric arrangement of the morainic hills with the concave side toward the Galiano Glacier. These morainic hillocks extended down to the shore of Yakutat Bay (PL XXXI); and that they also extended out under its waters was indicated by the pronounced shallowness of the water off shore along this coast, which caused icebergs to be stranded in far larger numbers, and to a greater distance out from the coast than is common along the shores of the bay. The moraine hillocks were found not only in the area of the former alluvial fan, but extended even to a distance of two or three miles to the south of Russell's photographic site. In this outer portion we have no photographic proof of the condition in 1890, but the close resemblance of the low, barren, conical hills, ridges and groups of morainic hills in this area to those already described leads us to class them together as to cause. This conclusion is further supported by the presence here of great numbers of tree trunks which littered the surface. Some of these, especially in the depressions near the coast, were doubtless brought and perhaps destroyed by the earthquake water wave which swept this coast in September, 1899; but others, resting above the line of driftwood that this wave brought, and even embedded in the moraine of the hillocks, cannot have been brought by the earthquake wave.
We feel confident, therefore, that some great change had occurred here which had de-